Select another Impact Assessment Method
Back to EPS default 2000

Characterisation Method Information
Characterisation Method Name:
PM10 impact on YOLL
Version:
1999
Date Completed:
1999
Principal Method Name:
EPS: acute, chronic effects and global warming pathway
Method Description:
YOLL stands for Year of Lost Life.


Model 1, pathways via acute effects

The characterisation factor is determined by the empirical method.

Total category indicator value in system considered

The first studies that were made on correlation between concentration of particles in ambient air and excess mortality were made on a day to day basis. Daily statistics were compared. Many studies were performed, with varying results. Typical values found of elasticity were in the range of 0.63 to 1.30 excess cases per 1000 deaths and mg/m 3 PM10 concentration with an average of 0.96 (Rosendahl, 1998).
When estimating the social costs of these effects the problem arises to estimate if this was a one-day premature death or more. When looking at the age statistics it shows that the increase of mortality mainly concerns elderly above 65 years. Rosendahl, (1998) concludes that there is no information about to which extent life shortening takes place, but if the average life time is in the order of 75 years, it seems reasonable to assume (applying the precautionary principle) that the years of lost life is less than 5 years, with an average reduction of 2.5 and an uncertainty range of 0.1-10 years. Rabl (1997) assumes that the mean reduction of life expectancy is 0.5 years corresponding to 4.68E-06
YOLL per person per mg/m 3 per year. In view of the more severe chronic impacts (2.6 ×E-04 YOLL per mg/m 3 , year and person, (see below)) the uncertainty does not seem to be crucial for the estimation of the overall effect. Considering the pattern of variation in particle concentrations it seems as even Rabl’s estimation is conservative, why this figure will be used in the modelling of characterisation factors below. The concentrations in various parts of the world vary, and few estimations have been made on a regional basis.

Typical values in Tokyo are in the order of 50 mg/m 3 for TSP. For the non-OECD (Organisation for Economic Co-operation and Development) world,
TSP concentrations are considerably higher than in OECD countries.UNEP/WHO (United Nations Environmental Program/World Health Organisation) made a study in 20 megacities (1992) and found concentration levels.

The concentration ratio between the non-OECD and OECD cities is about 4.5. As 50% of the population in non OECD countries live in urban rural areas and as more than 80 % of
the population in OECD countries live in urban areas, the ratio ought to be a little less in terms of exposure to the entire population. Assuming that the ratio also is relevant for PM10, that there is negligible exposure on he countryside and as about 1.2 billion live in the OECD world and 4.1 outside, the population weighted average concentration will be about 46 mg/m 3 , causing 1.14×10 6 YOLLs per year.

Contribution to total category indicators value from a flow unit

The global emission of PM10 contributing to the population exposure is not known, but
attempt have been made in some areas to estimate emissions of particles and gases that form particles in the atmosphere.

When comparing estimations of particulate emissions from different regions it is apparent that the knowledge about which emissions that really occur is insufficient, and that the use of official figures of known emissions ill give results that are too low. The best figures seem to come from the US.
To estimate the global emission of PM10 from technical processes an indirect method will
be used. This assumes that SO2 emissions are fairly well known as well as the average
concentration of SO2 and TSP in some of the worlds megacities and that the per capita
emission of particles is representative for the rest of the worlds urban areas. It also
assumes that the ratio of total yearly emission in megacities is the same as the ratio of yearly averages of concentrations in ambient air. The rational for this is that the sink processes are not fast enough to considerably decrease the amount of SO2 and TSP (total suspended particular matter) that is emitted from technical processes and that megacities are big enough to allow good mixing of emitted SO2 and TSP. The residence time for SO2 is in the order of several days and the wind velocity is in the order of meters per second, bringing emitted SO2 out of the area in a few hours. Some of the emitted TSP particles may be deposited giving a lower TSP-concentration than expected from the emission ratios. In table 9.5 below it can be seen that in cities where the knowledge of the air pollution situation is good, such as in
New York and Los Angeles, the TSP/SO2 ratio are approximately the same for emissions
and concentrations in ambient air, while they are very different in Bangkok, Beijing and Bombay. In Beijing, a large contribution from soil can explain some of the difference. In Bombay, there may also be diffuse emissions, but the difference may also be explained by the separation of the various source areas. The average per capita emission is 50 kg per year, when Beijing and Bombay are excluded. Using a typical TSP to PM10 ratio of 2, we obtain a total emission from the worlds 3 billion urban inhabitants of 74 million tons per year. Assuming the rural emissions from technical processes is about half of the urban (excluding industrial processes but including domestic and traffic) we obtain a total global PM10 emission of about 100 million tons per year.

When calculating the contribution to the total YOLLs estimated above it is however
necessary to consider the secondary particles formed from SO2, NOx and VOC emissions. According to USEPA cited by Wilson and Spengler (1996) PM10 consists of about 1/3 ammonium sulphates + nitrates, 1/3 of organic substances and 1/3 of minerals. Primary particles are therefore estimated to contribute with 2/3 as an average to the YOLLs estimated above. If one looks at the situation in rural areas the contribution is much smaller, but as a population exposure weighted average the 2/3-factor may be relevant for a large part of the world. The contribution to the YOLLs will therefore be 2/3×10 -11 per kg of PM10.

Calculation of pathway specific characterisation factor

1.14×10 6 YOLLs per year *2/3×10 -11 per kg of PM10 = 7.60 ×10 -6 YOLL/kg of PM10

Model 2, pathways via chronic

effects, such as cancer and decreased lung capacity

The characterisation factor is determined by an empirical method.

Total category indicator value in system considered

The global population weighted average concentration was determined in model 1 to 46
mg/m 3 . The risk, calculated by Rabl (1997) on the basis of results from Dockery and
Pope, is 2.61E-04 YOLL per person per year per mg/m 3 . The total impact on the globe
will thus be 6.34E+07 YOLLs per year.

Contribution to category indicators value from a flow unit

The same contribution apply as for model 1, i.e. 2/3E-11 per kg of PM10.

Calculation of pathway specific characterisation factor

6.34E+07 YOLLs per year *2/3E-11 per kg of PM10 = 4.23E-04 YOLL/kg of PM10

Model 3, global warming pathway

The characterisation factor is determined by an equivalency method using CO2 as a
reference.

Equivalency factor

The radiative forcing from CO2 is 1.5 W/m 2 (IPCC, 1994) and from tropospheric aerosols
–0.9 W/m 2 . The radiative forcing, F, is as a first approximation proportional to the global warming potential (GWP) and the global emission of a substance. The global emission is in turn proportional to the global average concentration C, divided by its average residence time, T. Thus
F = K*GWP*C/T or
GWP1*C1/(T1*F1) = GWP2*C2/(T2*F2)
The global average concentrations of CO2 and PM10 are 712 mg/m 3 and about 0.01 mg/m 3 respectively. The residence time is about 100 and 0.02 years respectively. The GWP for CO2 is 1. The GWP for PM10 may thus be derived as 712*(0.02*(-0.9))/(100*1.5*0.01) = – 8.54 relative to CO2.

Calculation of pathway specific characterisation factor

The characterisation factor will be –8.54* 7.93E-07 = -6.77E-06 YOLL/kg PM10, where
7.93E-07 is the added characterisation factors of four pathways for CO2’s impacts on
YOLL.

Calculation of the characterisation factor

The characterisation factor will be 7.60E-06 + 4.23E-04 - 6.77E-06 = 4.24E-04 YOLL/kg PM10

Literature Reference:
1. Rosendahl, K.,E., (1998) Health effects and social costs of particulate pollution - a case study for Oslo, Environmental modelling and assessment 3 (1998) 47 -61. 2. Rabl, A., (1997) “Quantifying the benefits of air pollution control: the Interpretation of Exposure-response Functions for Mortality”, Proceedings from RISK 97, Amsterdam , October 21 – 24 1997, organised by RIVM. 3. UNEP/WMO, “Urban air pollution in megacities of the world”, Blackwell Publishers, Oxford, 1992. 4. Wilson, R. and Spengler, J., “particles in Our Air: Concentration and health effects”, Harvard University Press, 1996, Harvard School of Public Health. 5. IPPC, “The 1994 Report of the Scientific Assessment Working Group of IPCC. Summary for policymakers”, WMO and UNEP, 1994
Methodological Range:
Particles become airborne through two types of processes: dispersion and condensation. Dispersion aerosols consist of comparatively large particles: from a few microns to several hundred microns, while condensation aerosols consist of particles from 0.001 to a few microns. The residence time of particles in air strongly depends of its size. A 10 mm particle has a settling velocity hundred times that of a 1 mm particle. The residence time is also influenced by condensation of water. During rain, particles are either trapped in the clouds when water vapour condense or by the falling drops through impaction. Particles between 0.1 and 1 microns are effective condensation nuclei, but are not efficiently caught by falling raindrops. For this, the size has to be larger than about 2 mm. The processes of emissions and depositions tend to stabilise the particle size distribution in air into a two-peak pattern. The fine particles often called the accommodation mode, are normally less than 2.5 mm and have a residence time in air in the order of several days. The large particles stay airborne during minutes to hours, and those found in air are normally of local origin. The emission of large particles depends heavily on humidity and wind velocity. In terms of mass concentration large particles may dominate occasionally while small particles have more stable concentrations and are considered to have a more severe impact on the environment such as human health effects after inhalation and effects due to soiling when deposited on surfaces. Sources of particles are widespread and frequent. Energy production, traffic, agriculture and various industrial activities contribute. Considering the location of sources to urban areas and the dispersion patterns, most of the exposure and effects on humans are likely to occur in the urban area where the emission occur. It would thus be possible to make characterisation models for each urban area without having to allocate effects to trans-boundary flows. For this characterisation model, we chose global system borders. The time period investigated is the year 1990.
Notes:

Existing Characterisation Factors of PM10 impact on YOLL
Characterisation Parameter Category Indicator Impact Indication Principle Aspect Substance Quantity Unit Notes
CFactor YOLL EPS/2000
Type = Emission
Direction = Output
Media = Air
Geography = *
PM10 4.24E-04 P yr/kg 3 pathways